Optical Exhaust Gas Temperature (EGT) Sensor and Instrumentation for

Gas Turbine Engines

Mehrdad Pakmehr, Joannes Costa, George Lu

optoXense, Inc.
San Ramon, CA 94583, USA
mp@optoxense.com

Alireza Behbahani
Air Force Research Laboratory
1950 Fifth St, Wright-Patterson AFB, OH, USA
alireza.behbahani@us.af.mil

ABSTRACT
This goal is to develop an inexpensive, robust, durable, and networkable EGT sensor and instrumentation for
operation in the extreme temperature environments of aircraft gas turbine engines. The next generation of adaptive
engines require low cost/weight and more robust instrumentation systems, controls, and PHM capabilities. EGT
measurement in gas turbine engines is considered a key parameter for optimizing fuel economy and PHM
systems. Currently, direct EGT measurements made on turbine modules are limited due to the ultra high
temperature (UHT) environment. Accurate EGT measurements and durable sensors are of specific importance
for high performance military turbine engines. A new FBG-based EGT sensor system is presented which involves
fiber optic sensing technology developed for operation in UHT conditions using sapphire fibers instead of the
silica fibers to raise the operating temperatures to 1600°C. Instrumentation development process includes: (a)
producing a minimally intrusive EGT sensor based on silica fibers but with UHT housing materials to account for
actual deployment conditions, (b) demonstrating a development path to prove viability of FBG temperature
sensing with sapphire fibers, (c) conducting laboratory tests and verification, and (d) developing a plan for engine
test cell experiments. Sensor ruggedness, precision, accuracy, and frequency response in a representative
environment will be studied. This optical EGT sensing system enables robust sensing in both legacy and next
generation engines used on military and commercial aircraft.

Key Words:

Sensor, Engine Instrumentation, Exhaust gas temperature (EGT), Turbine engine, control, PHM, Ultra high
temperature (UHT), Fiber optics sensing, Fiber Bragg grating (FBG)

1.0 INTRODUCTION
The exhaust gas temperature (EGT) is defined as the gas temperature at the exit of the turbine; the sensors used to
measure this parameter are considered the most vulnerable elements of the entire turbine engine instrumentation.
EGT measurement is needed for optimizing fuel economy, diagnosis, and prognosis. The reason is that turbine
blade temperature is a good indicator for normal life consumption of that blade [1]. Currently, direct sensor
measurements made on turbine modules are limited due to the ultra high temperature (UHT) environment. EGT

STO-MP-AVT-306 18 - 1
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

sensors located downstream from the highest temperature sections provide a means to roughly infer the
temperatures seen by the turbine blades/disks. But these sensors, which themselves are subject to frequent failures,
provide a fairly inaccurate indication of the actual metal temperature profiles. Accurate EGT measurements with
durable sensors are of specific importance for high performance military turbine engines where the margin between
hot section operating conditions and material limitations is shrinking [1].

EGT and other turbine temperature sensors are susceptible to degradation due to high temperature oxidation,
erosion and contaminant intrusion into probes and wiring harnesses. Thermocouples are easily affected by noise,
electromagnetic interference, and/or other environmental factors. Military and commercial field experience
indicates that gas path thermocouple removals affect aircraft availability and add maintenance time. Also, the
adaptive engines of the future are driving the control system to outperform legacy design, and cause higher
temperatures.

Existing technologies for measuring EGT typically implement high temperature capability with thermocouples
but extension to higher temperatures is not guaranteed. Other technologies that have been investigated include thin
film thermocouples, pyrometers, spectroscopy, and radioactive isotope-based sensors [1]. They are not mature,
accurate nor cost-effective for engine implementation. In addition to being not mature, multicolor pyrometers are
also complex, emissivity dependent, and expensive [1]. With alternate technologies that use fiber optic technology
[2][3][4][5][6][7] to measure exhaust gas effects, measurement of significantly higher temperatures is possible for
EGT sensing.

Fiber optic and laser-based sensing systems [2][3][4][5][6][7] are envisioned to be used for EGT measurement
and to accurately assess the condition and life usage of the turbine engine hot section, on a blade-by-blade basis.
To understand the extreme environmental conditions experienced in turbine engines, there is a need for robust
sensors to obtain high precision EGT measurements. The main objective of this work is to develop a new
technology for reliable EGT measurement using optical fibers in future high performance turbine engines. The
control and health management of modern turbine engines depends on sensing a wide variety of quantities
throughout the engine, (e.g., temperatures, pressures, and vibration, etc.) with different redundancy, reliability, and
accuracy requirements; reliable EGT measurement is one of the challenges for gas turbine engine control and
health management improvement.

An optical sensor system extends gas temperature measurement capability in turbine engines [8] beyond the
present generation of EGT sensing technologies. The sensing element which consists of fiber Bragg gratings
(FBGs) embedded inside a sapphire lightguide is capable of operating in UHT conditions. The sensor generates
an optical signal as a function of temperature. An optically averaged EGT measurement system can be developed
by combining the optical signals from multiple individual sensing probes at a single detector assembly.

Military turbine engines require the following specifications for EGT sensing system:

(a) Best entitlement for accuracy at higher than experience range 9~12°C (12°C temp margin ~1% thrust
margin) allotment in redline stack is required;

(b) It should survive the harsh environment while maintaining reliability, accuracy, ruggedness, and minimum
size/weight/power;

(c) As aircraft turbine engines continue to push the envelope on material capabilities, it is important to create
the capability of sensing close to the material limits for the operational turbine system;

18 - 2 STO-MP-AVT-306
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

(d) It should be able to survive the expected life of the engine between overhauls and measure temperatures
in excess of 1600°C (Life: 2000 Engine Flight Hours (EFH), immersive - 4000 EFH, non-immersive);

(e) The accuracy of the sensing system should be 0.5% of full scale and stable over the life of the engine;

(f) Air temperature measurements in the exhaust gases must be taken outside of the wall boundary layer;

(g) It should be ruggedized for installation in production aircraft.

To address these needs, a new FBG based EGT sensor system is needed, which involves optical fiber sensing
technology developed for operation in UHT conditions using sapphire fibers instead of the more common silica
fibers. This will raise the operating temperatures from ~1000°C to ~1600°C. The development is a three-step
process. First, since sapphire based optical sensors have yet to be reduced to practice, we will produce a minimally
intrusive EGT sensor design based on silica fibers but with UHT housing materials, to account for actual
deployment conditions. Second, as FBG sensors inherently perform better under single mode conditions while
typical sapphire fibers are multimode conduits, as well as the fact sapphire fiber segments need to be kept short
for optomechanical and cost reasons, we will demonstrate a development path to prove FBG temperature sensors
embedded in sapphire fibers is indeed viable. Third, in addition to laboratory tests and verification, optoXense
will develop a practical plan for engine test cell experiments, and a development path for commercial application
of the technology. We will also plan to demonstrate the sensor ruggedness, precision, accuracy, and frequency
response in a UHT environment at representative engine test facilities.

2.0 OPTICAL EGT SENSOR DEVELOPMENT

2.1 Technology tradeoff studies

The requirements and specifications of the sensor include: (a) sensor size, material, and specifications for design
and fabrication; (b) optical sensors positioning and installation process; (c) a plan for test and verification of the
sensor system at an engine test cell.

The tradeoff study for various high temperature optical fibers and protective materials is performed to select the
fiber type, high temperature protective material type, and also various fiber coating options. The following tables
provide information on fiber type, high temperature protective material type, and also various fiber coating options.
Table 1 is a preliminary comparison of two high temperature fibers which can be used for the tradeoff study.

Fiber optic sensors need to interact with the environment if one wants to achieve a certain level of sensitivity to
the measured parameter of interest. They should ideally also survive the lifetime of the monitored system.
Particular care needs to be taken to provide the sensor and the fiber leads with adequate packaging and protection.
A fiber optic sensor probe typically consists of different components and materials, including an optical fiber,
special coating materials or chemical reagents, dedicated adhesives, filter or reflective elements, etc. This
complicates the assessment of the reliability as all the different sub-parts of the sensor probe should be thoroughly
evaluated. Additionally, fiber sensors are often used in harsh and very specific conditions.

STO-MP-AVT-306 18 - 3
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

Table 1: Comparison of ultra high temperature fiber sensor technologies

Optical Fiber
Advantages Disadvantages Comments
Technology
Silica fiber - Flexible fiber supporting Limited to ~1000°C This is the most mature high
(with 193 nm or 800 multiple sensor points. temperature fiber sensor
nm laser light) - Could be excellent for lead-in technology. optoXense staff has
fiber attached to sapphire probe demonstrated technology working
for highest temperatures to 1000°C for fiber alone.
Sapphire fiber - Sapphire fiber has a melting When different modes Stephen Mihailov has fabricated
point of ~2050°C. are excited there will be multiplexed FBGs in sapphire fiber
-Demonstrated to ~1400°C an extra phase shifting using an 800 nm femtosecond pulse
long term for the modulated light, laser [9][10].
-Multimode fiber but adaptable which will cause
for the proposed sensor system measurement error.

The durability and reliability in long term service is not guaranteed by just conducting acceptance tests. In this
respect, more research is required to understand failure mechanisms and for estimating those parameters that are
essential to long term functionality. Interfacial parameters such as normal stress, friction, and decohesion energy
of fiber/coating or coating/embedding matrix are important for the stress transfer to the sensor. There is no lifetime
model available neither for embedded fiber sensors under load, nor for specially coated fibers that face long term
exposures to a variety of chemicals. High temperature is especially an issue for fiber optic sensing in gas turbine
engine.

Table 2: Laser writing procedures for optical fibers and FBGs

Option Annealing Description Comments Ref.

Writing with a typically at 248 nm
FBG reflectivity can decrease above
(KrF) or 193 nm (ArF), but also IR at
No 250°C with complete erasure above [11][12]
Type I-UV 800 nm with hydrogen loading and
800°C.
lower laser intensities
FBG can be regenerated so that it is
Yes [13]
stable at 1000°C
Pulsed writing results in period
Writing with a high power pulsed UV
Type II-UV compaction – Effect on strength vs [14]
laser
Type I to be evaluated
Writing with an ultrafast high power
UHT-IR Sensor theoretically robust to 1750°C [14]
pulsed IR laser

FBG reliability and durability highly depend on the manufacturing procedure in which there can be considerable
variation by different developers and manufacturers. Those that are manufactured optimally will be much stronger
than those that are not. Nevertheless, there are tradeoffs. These tradeoffs could be evaluated/tested during our
sensor design and analysis. Table 2 and Table 3 show some preliminary considerations for design and analysis
based on the fiber manufacturing methods and the various available technology options.

18 - 4 STO-MP-AVT-306
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

Table 3: Fiber coatings procedures

Option Option Name Description Comments

Fiber can be damaged during stripping, recoat is not as
Strip buffer layer, inscribe
robust as original coating. In the early days the buffer
Option 1 Strip & recoat FBG, recoat (Chemical vs.
layer was removed mechanically with microcracks
mechanical stripping)
resulting thus considerably weakening the FBG region.
FBG inscribed through Reflectivity is generally lower than after writing without
Option 2 Write through polyimide buffer layer buffer layer. Not applicable to thicker acrylate buffer
without removing it layers
FBG inscribed as fiber
Fiber is moving, so inscription must be rapid; precision
Option 3 Draw tower being drawn before buffer
not as good as Option 1.
layer coating applied

Whenever a drift of the EGT sensor data is observed the question has to be answered if the aging of the used
sensors, or the instrument causes the drift. Proper sensor applications do not only require the monitoring of the
sensor signals per se, but also condition monitoring of the environment, the sensors, and the sensor system. Self-
diagnosis allows to discriminate between a possible permanent degradation of the sensor, a failure of the
monitoring system, or the optical fibers, or a reversible short term loss in signal reliability due to environmental
influences.

2.2 Design, fabricate, and characterize proof-of-concept optical EGT sensor

The goal is to develop a durable and high precision sensor suite. The sensor suite will consist of optical EGT sensor
elements. One of the important goals is to develop high precision durable EGT sensor system that can operate in
turbine engine UHT sections.

The design process focuses on functionality in UHT environment, durability, and robustness of the sensor system.
Moreover, our approach adds multiplexed sensors and the ability to: (a) develop a nonintrusive installation process
for the measurement system, (b) have a high precision and robust sensing system, and (c) make measurements at
multiple points in UHT gas turbine engine environments without extensive harnessing requirements.

Figure 1 and Figure 2 provide simplified schematics of the developed EGT sensor probe. This is a proof-of-concept
optical sensor system.

UHT fiber optic temperature probe design

Figure 1 shows a minimally intrusive EGT sensor core design based on silica fibers, though we will be using UHT
probe housing materials to account for actual deployment conditions for the mechanics.

STO-MP-AVT-306 18 - 5
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

Figure 1: UHT fiber optic probe body design.

The photonics goal is to push the standard silica based FBG sensors to their limits and beyond. The mechanical
goal is to minimize physical space impact for the UHT section of the probe, while providing enough probe body
for handling, mechanical mounting, and thermal dissipation outside of the UHT zone, and to confirm the structure
will survive the anticipated 1600C operating conditions. We emphasize the “core” designation, because in
practice this element may not be deployed as is, but rather inserted into appropriate sensor acceptance ports
designed for production processes and maintenance routines.

The main features of the sensor core are as follows:

• The design aspect ratio of the probe body is 52:1, based on its largest diameter.
• The probe housing material limits (ceramic ~2200C) will suffice for this exercise.
• The angled end of the fiber is to diminish the effect of spurious reflected signals and resonances which
will impact FBG signal response analyses.
• The fiber end containing the sensor array is free to experience thermal expansion and contraction, thereby
decoupling mechanical strain effects from the sensor array’s thermal response.
Limited distance sapphire FBG sensor studies
As noted previously, there are several concerns related to the optical viability of generating well defined FBG
response signals. Our response is to go back to fundamentals and initiate a different path for reduction to practice,
especially with the specific application in mind – minimally intrusive UHT sensors for turbine engines, where a
significant propagation length does not yield an appreciable advantage since the sensor portions need to be small.
The basic experimental setup is shown in Figure 2.

18 - 6 STO-MP-AVT-306
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

Figure 2: Sapphire FBG parameter effect setup.

Although this is a room temperature evaluation, we will still employ UHT FBG elements within the sapphire fiber,
to avoid inconsistencies in future experiments based on this initial work.

Most sapphire fibers do not have a round cross section, but rather a rounded hexagonal shape. The material is
uniaxial. And the surfaces are not perfectly smooth. Therefore, we start by introducing collimated light into the
sapphire fiber, which will minimize the generation of large angle light rays and minimize scattering losses. Note
that “collimated” is a relative term. Given the wavelength, aperture size, beam size, and the associated optical
collimation component performance limits based on these parameters, the light rays incident on, and propagating
through, the sapphire fiber will not all be perfectly parallel. To address this, the input gap distance L0 between the
collimator output and fiber input will be adjusted for optimum response in terms of sensor signal selectivity and
minimal noise.

We will then fix the sensor response wavelength (e.g. 1540 and 1550nm) and examine the primary parameters –
diameter D and propagation distance Lx – against FBG signal response characteristics. After data for optimum L0
has been documented, response degradation vs. L0 can be examined.

2.3 Evaluation and verification of the EGT sensor in the laboratory environment
For the first step of calibration, we concentrate on the static calibration. Since the device has to survive the tests,
and for this stage of the development the FBG elements are hosted in silica, the low to midrange temperature zone
(25~800C) will be examined in detail first. After an initial scan of the overall FBG response vs. temperature plot,
we will further define additional temperature regions with higher resolution increments to highlight any
nonlinearities or perturbation features in the calibration plot. Each temperature point will require a thermal
stabilization timeframe for the oven, which in turn will anneal the elements at those particular temperature points,
and after further time to allow settling of the FBG response this will also yield short term drift data. Before moving
onto the higher temperature regions (>800C), a basic hysteresis evaluation will be performed by reversing the
test path towards lower temperature points. Then on the cycle back up, the high point of the midrange (i.e. 800C)

STO-MP-AVT-306 18 - 7
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

can be used for a longer term drift test. Finally, at the higher temperatures we expect the array to yield a linear
calibration response, and we further expect to observe the point at which the FBG signal levels start to decrease
due to the higher temperatures affecting the permanence of the FBG elements within their silica host. This
irreversible constraint, indigenous to this work, is why the majority of the sensor calibration and evaluation
processes needs to be demonstrated at the low to midrange temperature zone. Temperatures beyond the FBG
response limit will be used solely for sensor housing survival studies. We will conduct dynamic sensor response
studies (e.g. transient heating using a focused high power pulsed laser beam) and engine test cell evaluations in
near future.

Figure 3: UHT EGT fiberoptic probe response setup for static calibration.

High temperature oven tests: Temperature tests will be done using a small cavity UHT oven, as shown in Figure
3. As example of such a UHT unit is the FIB-THM500 fiber heater by Micropyretics Heaters International. Static
incremental temperature points will be defined to establish a wavelength vs. temperature calibration profile, to
demonstrate overall temperature performance. More specific results will be mechanical survival of the housing,
and any impact on data integrity. We expect to witness the failure of the FBG gratings due to the limited operating
temperature of their silica host.

Frequency response test setup design: An experimental test setup is developed and shown in Figure 4 for testing
the frequency response of the proposed EGT sensing system. The heat source is an 830nm diode laser capable of
2W maximum optical power (e.g. ThorLabs LB830ME2W), and a fast laser driver (e.g. Highland D200), to get
up to 100 kHz laser modulation rates with adjustable duty cycles. The laser emitter is typically 1 µm x 100 µm,
and can be imaged onto the FBG to a new extent of 10 µm x 1000 µm, thus achieving up to 20 kW/cm2 peak
intensities. This should be enough to yield good S/N during FBG temperature response tests.

18 - 8 STO-MP-AVT-306
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

Figure 4: High frequency, high intensity optical test setup to evaluate the thermal response time of an
FBG temperature sensor.

3.0 CONCLUSION
Currently, direct EGT measurements made on turbine modules are limited due to the ultra high temperature (UHT)
environment. Accurate EGT measurements and durable sensors are of specific importance for high performance
military turbine engines. A new FBG-based EGT sensor system development is presented here which involves
fiber optic sensing technology developed for operation in UHT conditions using sapphire fibers instead of the
silica fibers to raise the operating temperatures to 1600°C. This work produces a minimally intrusive EGT sensor
based on silica fibers but with UHT housing materials to account for actual deployment conditions, as the first step
to demonstrate viability of FBG temperature sensing with sapphire fibers. This is an ongoing work, currently we
plan to conduct laboratory tests and verification, and then to conduct engine test cell experiments. Sensor
ruggedness, precision, accuracy, and frequency response in a representative environment will be studied. This
optical EGT sensing system enables robust sensing in both legacy and next generation engines used on military
and commercial aircraft. This technology is not only applicable for military engines, but also for commercial
aircraft and automobile engines.

STO-MP-AVT-306 18 - 9
 Optical EGT Sensor and Instrumnetation for Gas Turbine Engines

REFRENCES
[1] A. Von Moll, A. Behbahani, GC Fralick, JD Wrbanek, GW Hunter, “A Review of Exhaust Gas Temperature
Sensing Techniques for Modern Turbine Engine Controls,” 50th AIAA JPC, Cleveland, OH, July 2014.

[2] Tregay, G., Calabrese, P., Finney, M., Stukey, K., “Durable Fiber-Optic Sensor for Gas Temperature
Measurement in the Hot Section of Turbine Engines,” SPIE Vol. 2295 Fly-by-Light, 1994.

[3] M. Willsch, T. Bosselmann, P. Flohr, et. al, "Design of Fiber Optical High Temperature Sensors for Gas
Turbine Monitoring," Proc. SPIE 7503, 20th Intern. Conf. on Optical Fibre Sensors, 75037R, Oct 5, 2009.

[4] Hua Xia, Doug Byrd, Sachin Dekate, and Boon Lee, "High-Density Fiber Optical Sensor and
Instrumentation for Gas Turbine Operation Condition Monitoring," Journal of Sensors, Vol. 2013, 2013.

[5] D. Grobnic, S. J. Mihailov, C.W. Smelser, and H. Ding, "Sapphire Fiber Bragg Grating Sensor made using
Femosecond Laser Radiation for Ultrahigh Temperature Applications," IEEE Photonics Technology Letters,
Vol 16, No. 11, pp. 2505-2507, 2004.

[6] A. Wang, H. Xiao, J. Wang, Z. Wang, W. Zhao, R. G. May, "Self-Calibrated Interferometric-Intensity-Based

Optical Fiber Sensors,". Journal of Lightwave Technology, Vol 19, No. 10, pp. 1495-1501, 2001.

[7] M. Gottlieb and G. B. Brandt, "Fiber-Optic Temperature Sensor Based on Internally Generated Thermal
Radiation," Applied Optics, No. 19, Vol. 20, pp. 3408-3414, 1981.

[8] M. Pakmehr, J. Costa, A. Behbahani, et. al, “A Review of Fiber Optic Networks for Turbine Engine
Instrumentation Channel: Control, PHM, and Test Cell Applications,” 50th AIAA Joint Propulsion
Conference, Cleveland, OH, 2014.

[9] S.J. Mihailov, D. Grobnic, and C.W. Smelser, High temperature multiparameter sensor based on sapphire
fiber Bragg gratings,” Optics Letters 35(16), 2810-2812, August 15, 2010

[10] S.J. Mihailov, “Fiber Bragg Grating Sensors for Harsh Environments,” Sensors 12, 1898-1918, 2012.

[11] C. W. Smelser, D. Grobnic, S. J. Mihailov, “High Temperature Stable Fiber Bragg Gratings in hydrogen
loaded all-silica core fiber,” OSA/FiO/LS/AO/AIOM/COSI/LM/SRS 2009, Paper FTuD7.

[12] K. O. Hill and G. Meltz, “Fiber Bragg Grating Technology: Fundamentals and Overview,” JOURNAL OF
LIGHTWAVE TECHNOLOGY, VOL. 15, NO. 8, 1263-76, AUGUST 1997.

[13] J. Canning, “Regenerated Gratings for Optical Sensing in Harsh Environments,” Advanced Photonics
Congress, Paper BTu3E.3, 2012 OSA

[14] J. L. Archambault et al, “Type II FBGs,” Electron. Lett., Vol. 29, 1993, p. 453.

18 - 10 STO-MP-AVT-306